Skip to Main Content
With the presentation of massive remotely sensed data, it is one of the biggest challenges how to process and analyze these data as soon as possible. Owing to Grid conformity heterogeneous computing sources, a Grid environment is built for the processing of remotely sensed images. In this study, CSF4 is taken as meta-scheduler in the collective layer in such a network environment. The message transmission is implemented by a protocol defined by a Grid middleware GRAM (Globus Resource Allocation Manager). SGE, LSF, and OpenPBS are used in the fabric layer of the Grid environment. As an example of remotely sensed image processing in the application layer, image smooth processing is achieved under the MPICH-G2 programming model. The relationship between the node number and time-consuming are analyzed. And the efficiency is shown by comparison between the parallel and serial processing under different node numbers and image sizes.